The key objective of the proposed research project is to identify some of the factors that determine the perceived strength of inductive categorical arguments such as {All Dogs Have Substance X, All Pigs Have Substance X, All Animals Have Substance X}. The hypothesis-assessment model that drives this research predicts the strength of the preceding argument to be function of 1) the number of premise categories that instantiate the conclusion; 2) the extent to which the premise information must be generalized, and; 3) the accessibility of competing conclusion categories. A set of experiments is proposed to test the model's ability to account for variability in perceived argument strength under various conditions. Both task and argument parameters are manipulated. The behavior of subjects on inductive reasoning tasks is widely recognized as being relevant to the judgments of medical doctors and other health professionals (e.g., Baron, 1988; Chapman & Chapman, 1967, 1969; Dawes, 1988; Eddy, 1982; Elstein, Holzman, Ravitch, Metheny, Holmes, Hoppe, Rothert, & Rovner, 1985; Jaccard & Wan, 1993; Lehman, Lempert, & Nisbett, 1988; Medin, Altom, Edelson, & Freko; Schwartz, 1986). The formulation of diagnoses, prognoses, and treatment plans, for example, always requires inductive generalizations of past evidence to current cases. Unearthing those factors that make inductive inferences appear stronger, especially factors that can lead to nonnormative conclusions, would be of critical importance in the training of health care professionals. Further, the use of effective argument forms could significantly affect the physician's influence on the behavior of people who are at risk for AIDS and other STD's (Jaccard, Helbig, Wan, & Gutman, 1990; Brown, DiClemente, & Park, 1992; Norris & Ford, 1992; Rosenthal, Hall, & Moore, 1992; Wilson, Jaccard, Endias, & Minkoff, 1993).